Article Text

Download PDFPDF

Cost-value Analysis in Health Care: Making Sense out of QALYs
  1. John McMillan
  1. Ethox and University College, University of Oxford

    Statistics from

    Request Permissions

    If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.

    Eric Nord, Cambridge, Cambridge University Press, 1999, 175 pages, £35 (hb) £11.95 (pb).

    Eric Nord's book is required reading for all those interested in resource allocation. It is largely a book on health economics, but the importance of the issues discussed and the clarity of this book mean that it is relevant to all those involved in resource allocation.

    One of the more common objections to QALYs (Quality Adjusted Life Years) is that they focus on maximising the benefit produced by health care without paying attention to other factors relevant to allocation. One of the basic theses of Nord's book is that there is something right about this line of criticism and he sets out to correct cost-effectiveness studies by building into them societal concerns for the way in which health care resources are distributed. In chapter four he argues that the standard approach to QALYs is inconsistent with two major societal concerns about fairness. First it violates the strength with which people wish to give priority to the severely ill over the less severely ill at the expense of the total amount of health produced. Second, QALYs overlook the societal concern to allow people to realise their potential for health, whether this is big or small. Reasons such as these mean that “ . . . to rank projects in terms of costs-per-QALY as often as not may tend to distort resource allocation decisions rather than to inform them” (page xx). He proposes that the utility values used in cost-effectiveness equations be generated from representative sampling of persons who have received treatment.

    One of the really interesting differences between Nord's approach and that of the QALY concerns the appropriateness of trade-offs between gains in quality of life and length of life. If QALYs are interpreted as measures of psychophysical quantities, then such a trade-off exists by definition, but when we follow Nord's suggestion and go for individual valuations it becomes an empirical question whether people really are prepared to make such quality/quantity trade-offs under all circumstances. He argues that there is evidence for individuals not being prepared to make such trade-offs when services lead to minor functional improvements and when patients face life-saving procedures.

    As Nord points out, most people who use QALYs want them to be applicable across the whole range of health care interventions, so for example they are used to look at the relative cost-per-QALY of hip replacements and heart transplants and conclusions are drawn about where money ought to be spent. QALYs focus on maximising health benefits to individuals and fairness seems to require more of us than that, QALYs need to be based upon the societal value of health care programmes. In chapter six Nord sets out “the person trade-off technique”, which he advocates for ascertaining societal preferences.

    Empirically this will work by researchers specifying the characteristics of outcomes of different kinds and asking a representative sample to express the number of people obtaining one kind of outcome that they would regard as equivalent to a given number of people obtaining another outcome.

    While I think this is an excellent book, and I recommend it highly, I have two worries about Nord's approach. First, when thinking about the crucial question of how we can best allocate health resources it is tempting to think that it must be done on the basis of a theory of allocation that can provide us with the definitive answer. Nord thinks there may be something inherently misleading about allocating treatments on the basis of QALYs alone (a very plausible belief). However, many people who actually make decisions about how resources are to be distributed use QALYs, but only as one tool in a process of deciding about what will be funded. A good allocation decision should involve careful deliberation about a number of relevant factors, for example, how urgent the treatment is, and whether the group suffers a disability also, as well as attention to cost-effectiveness. So Nord may be right to argue that the standard QALY is not sufficient for making the right decisions about allocation but I am optimistic that those making these decision do not make them on the basis of QALYs alone. Second, I am still concerned about what it is we are measuring when we discover individuals' preferences via the direct person trade-off technique. The assumption seems to be that there is nothing more to ascertaining fairness than getting an accurate picture of what people in general believe about various allocations. Arguing about fairness and justice is a complicated matter and it may well be that people have ill-considered or unjustifiable beliefs about what this amounts to. Furthermore there is no single theory of justice that everyone accepts. Once we assume we have a tool that incorporates societal concerns about fairness then presumably we have the magic key that will unlock resource allocation decisions for us. An advantage of the present situation is that if our health authority makes an allocation decision that we think is unfair or unjust we can apply pressure on them to explain why they think this a just decision. Inevitably when they do this they implicitly or explicitly use one conception of justice or another, but this is something for which they must take responsibility and for which they can accept criticism.

    Other content recommended for you